8/2/2019 stowell060301ab

1/4

JULIA STOWELL, STEPHEN LAUBACH, Bureau of Economic Geology, TheUniversity of Texas at Austin, TX; RANDALL MARRETT, Department of Geological

Sciences, The University of Texas at Austin, TX; JON OLSON and JON HOLDERDepartment of Petroleum and Geosystems Engineering, The University of Texas at

Austin, TX.

Understanding Fractured Carbonate Reservoirs

Our understanding of fractures in carbonate reservoirs is hindered by low data density,

which prevents effective fracture attribute mapping. Despite the advent of image logs andnew core-recovery techniques, fracture data are frequently incomplete and sometimesmisleading. Sparse sampling of large fractures remains unavoidable. From the interwell

region, information is confined to seismic identification of faults and layer curvature.Although there may be a link between layer curvature and fracture intensity in situations

where fractures develop during or after layer flexure, if fracturing predates flexure there

is no such relationship. Collection of meaningful, systematic data at the well bore andextrapolation into the interwell volume are significant challenges.

New techniques, developed initially in siliciclastic rocks and that use microstructures to

predict orientation, population systematics, and openness of macrofracture sets, havebeen applied to fractured carbonates. We have been able to collect microfracture data inseveral dolomite studies. Calibration of the microfracture populations with observed

macrofracture populations in outcrop analogs is under way. For example, we areconducting an investigation of interwell heterogeneity in Permian Clear Fork dolomitereservoirs found in the Permian Basin of West Texas and New Mexico. The approach is

to study the stratigraphy, fractures, and petrophysics of outcrop analogs (Clear Fork

exposures) in the Sierra Diablo Mountains, West Texas, and to apply the results tosubsurface Clear Fork reservoirs, principally the South Wasson Clear Fork reservoir. Weenvisage that both the techniques that are developed and the results will be applicable to awide range of fractured carbonate reservoirs.

Macrofractures in limestones and chalks show many characteristics similar to those of the

macrofractures observed in siliciclastic rocks (e.g., power-law aperture-sizedistributions), but populations of microfractures have proved more difficult to observe inthe samples we have studied. In the case of the Austin Chalk, the fracture intensity is so

low that even the microfractures have very low intensities on the scale of a thin section.Low fracture intensity limits the technique we have developed to fracture-quality

prediction and determination of fracture orientation. Scaling work in these cases requiresthe use of outcrop analogs. However, information gained from the fracture qualityassessment can help to steer outcrop-analog selection because structural diagenesis

histories of rocks at outcrop can be matched to those of subsurface rocks as closely aspossible.

For subsurface studies, we quantify fracture attributes and flow potential on a bed-by-bedbasis using samples from whole core, if available, and wireline-sidewall cores if whole

8/2/2019 stowell060301ab

2/4

core has not been taken. Horizontally orientated thin sections are taken in order to

establish the fracture orientation(s), crosscutting relationships, and structural diagenesisof each sample. Structural diagenesis encompasses the relative timing of cementation

and fracturing and the relative proportions of different cements that have precipitatedbefore, during, and after each fracturing event, termedpre-, syn- and postkinematic

cements, respectively. Because fracture systems in carbonate reservoirs are diverse inorigin, a new emphasis on structural diagenesis research is essential. All too often,unjustified assumptions are made about relationships between fractures and large-scale

structures, without reference to rock properties at the time of fracturing, as affected bydiagenetic history. Patterns of cementation within developing fractures have beenestablished through studies of siliciclastic rocks, allowing us to predict whether

macrofractures are likely to be open to fluid flow. The complexity of carbonatediagenesis makes this work a special challenge in carbonate reservoirs, but we see some

parallels in the basic pattern of fracture mineral fill. It is common for a synkinematiccement to line the fracture, with bridges of that cement extending across the fracture fromone wall to the other. The porosity of that fracture after this phase of cementation either

remains open (Fig. 1a) or becomes occluded by a postkinematic cement, either partly(Fig. 1b) or completely (Fig. 1c). Observations of fracture populations, however, indicate

that there is a fracture aperture-size control over the extent to which the synkinematiccement fills the fracture. Most microfractures are completely filled with synkinematiccement, whereas macrofractures may have some porosity after synkinematic cement

precipitation has ceased. The aperture size at and above which porosity may be preservedis termed the emergent threshold. Observations to date reveal that the emergent threshold

is approximately 1 cm in limestones but less than 1 mm in dolomites. The geochemicalcontrols over this phenomenon are to be modeled as part of a new project.

Subcritical crack growth is an important mechanism for natural fracture development.Boundary element geomechanical models have been developed that show that fracture

spacing and clustering are sensitive to the subcritical crack growth index, n, which is theexponent used to describe power-law dependence of crack velocity on stress intensity. Atesting rig has been developed for measuring the subcritical crack index in sedimentary

rocks. We have carried out subcritical crack index measurements on chalk and dolomitesamples. The next stage is to input the results into the geomechanical models in order to

predict two-dimensional fracture architecture on a bed-by-bed basis. The results of themodels may be compared with empirically derived models for fracture-populationattributes such as aperture, length, and spacing. This approach is intended to improve

assessment of mechanical and fracture stratigraphy and interwell fracture-patternprediction. The attributes of the whole fracture population, which impact the flow

properties of the system, may then be used to improve flow simulations. The combinationof these core-based techniques represents a holistic view of the fracture-rock system thatcould lead to a renaissance in carbonate fracture studies.

8/2/2019 stowell060301ab

3/4

(a)

(b)

(c)

Figure 1. Photomicrographs of fractures showing different degrees of mineral fill. (a) Fracture in Austin

Chalk outcrop. Synkinematic euhedral calcite crystals line the fracture, pointing into open pore space in the

fracture center. The left-hand end of the fracture is completely filled by this cement . (b) An early phase of

calcite cement lines this fracture, and a later phase of calcite (coarser pale grains) has precipitated over the

earlier calcite, forming mineral bridges. Postkinematic quartz partly infills the remaining pore space. This

1.2-mm-wide fracture is from the Upper Cretaceous Agua Nueva Formation from Southern Mexico. (c)

Outcrop of opening-mode fracture in the Pennsylvanian Marble Falls Limestone at Pedernales Falls State

Park. An early, pale-colored calcite cement lines and bridges the fracture. A later, dark cement

(composition not currently known) completely fills the fracture. Lower scale bar = 0.5 mm.

8/2/2019 stowell060301ab

4/4